GHINS, Michel. Scientific Realism and Laws of Nature: A Metaphysics of Causal Powers. (Springer, 2024, xviii + 211 pp.)Abstract
This is a critical analysis of Scientific Realism and Laws of Nature: A Metaphysics of Causal Powers, by Michel Ghins.
Keywords:
Ghins; Scientific Realism; Laws of Nature; Causal Powers; Metaphysics of Science
The book Scientific Realism and Laws of Nature: A Metaphysics of Causal Powers (henceforth, SRLN) has two main points: (1) a defense of scientific realism in terms of a bottom-up approach, as an alternative to the explanationist top-down account grounded on theoretical virtues; (2) a defense of a dualist metaphysics of nature according to which there are categorical properties and natural powers. Michel Ghins divides the book into five chapters. In chapter one, Ghins defends what he calls a synthetic conception of theories, i.e., the claim that theories consist in both propositions (syntactic part) and models (semantic part). Furthermore, he claims that models have two roles: they make true the laws of the theory and represent the structures of properties. In chapter two, he develops a reply to “the loss of reality objection” to models, i.e., the claim that models cannot represent any external entities. He also defends a correspondentist account of truth, grounded on the existence of truthmakers. Besides, Ghins analyzes Putnam’s “no miracle argument” and concludes that realism is a philosophical, not a scientific position. In chapter three, he develops his bottom-up strategy for justifying scientific realism and also addresses some possible objections usually made to realist approaches in general. In chapter four, he establishes a distinction between categorical and dispositional properties. Furthermore, he argues for the necessity of laws, in some sense that is distinct from the necessity of logical truths, and that distinguishes laws of nature from merely accidental true statements. In chapter five, Ghins vindicates a neo-Aristotelian dualist metaphysical approach of causal powers as a way for grounding his theory. Natural powers make scientific laws necessary in the special sense he claims, a kind of necessity that is stricter than the logical one.
Ghins’s defense is well articulated and it involves a very comprehensive project. He couples a defense of scientific realism with an underlying metaphysics of causal powers and nature that presents an account for his bottom-up strategy for grounding scientific knowledge, and our consequent approximation from truth inasmuch as theories are developed and replaced. Given the richness of Ghins’s book, and the many themes he addresses in it, it would require me to write a new book in order to address everything that is discussed. In this review, then, I will limit myself to present Ghins’s argumentation on the subjects I chose to discuss and, then, present some critical remarks, following the schematization of the two main points I highlighted in the beginning of this review.
1. Scientific Realism
SRLN proposes a new strategy for defending scientific realism, different from the more common contemporary approach involving the appeal to theoretical or supra-empirical virtues such as simplicity, explanatory power, fruitfulness, etc. This common way for grounding scientific realism is what Ghins calls “explanationism” or “top-down” approach. A classic example presented by Ghins is Putnam’s famous “no miracle argument”, according to which the predictive success of our best scientific theories (such as those involving DNA and electrons) would be miraculous if the entities they posit do not exist. The alternative, and simpler, explanation is to acknowledge that these theories are at least partially true and, hence, the entities they refer to actually do exist. Furthermore, Putnam (1978, p. 18) argues that this would be a scientific rather than a philosophical thesis. As Ghins (p. 59)1 points out, this argument recurs to abduction, or inference to the best explanation, as a means of supporting the truth of a hypothesis. Thus, if hypothesis H explains a fact better than any other competing hypothesis, then H is true (or more likely to be true). Ghins emphasizes that more than just the rationality of belief is at stake in this matter. The realist must demonstrate that it is unreasonable not to believe in the alternative account (i.e., the one that is not “the best explanation”), and it is actually impossible to compel someone to believe in unobservable entities under penalty of irrationality (p. 60; van Fraassen, 2017, p. 102).
Ghins introduces a distinction between two kinds of explanations, drawing an analogy to the distinction between valid and sound arguments. We know that a sound argument is a valid one with true premises. Analogously, we have a satisfactory explanation when it provides a good understanding or comprehension of the phenomena, but it may not necessarily be true. An example is Ptolemy’s astronomical model which explains the stable motions of the planets by means of the regular rotation of crystalline spheres. Although it offers a clear understanding, there are no such spheres, and thus, Ptolemy’s model is false. A true explanation is equivalent to a sound argument: it is a satisfactory explanation in which all assumptions are correct (pp. 60-1). I agree with Ghins that inference to the best explanation can, at most, make our account a satisfactory one, but it never reaches the point of making it true (p. 83); in other words, inference to the best explanation is not a truth-conducive procedure. He repeatedly calls our attention to the fact that abductive reasoning aids us in forming new hypotheses, but its value is purely heuristic (p. 61 [n. 24], 62, 83, 84, 90, 176). As Ghins (p. 67) also notes, we usually subject our abductive hypotheses to empirical testing before confirming them; so, ultimately, it is not abduction itself that is truth-conducive, but rather the empirical testing and, therefore, an inductive form of reasoning.
Ghins claims that “the reasons for believing in the reality of detected entities are analogous to those that justify our beliefs in the existence of ordinary things” (p. 72). He presents four conditions, called the OCIM conditions: Observation Condition, the Causality Condition, the Invariance Condition, and the Measurement Condition. The Observation Condition specifies that we need to detect (either directly or indirectly, using instruments such as microscopes or telescopes) a property in order to have good reasons to believe in its existence (p. 74). The Causality Condition requires that we detect a causal connection between a property that is not directly observed to properties that are directly observed, using reliable instruments, to justify belief in the existence of unobservable entities (p. 77). This condition raises a problem: how do we attest the reliability of these instruments of detection? Ghins claims that “[w]e can check the trustworthiness of a new instrument by comparing the measurements it delivers with those obtained in the same empirical domain by an instrument the accuracy of which has been previously tested” (p. 77). This comparison would serve as an inductive justification of instrument reliability. The Invariance Condition requires that repeated detections of a property, using distinct and independent empirical methods, yield approximately identical results in order to provide reasons for believing in the existence of the detected property (p. 79). Lastly, the Measurement Condition is applicable to the quantitative sciences and requires that a property be quantitatively measured by reliable instruments to justify belief in the existence of this detectable property (p. 79).
Ghins’s bottom-up method, therefore, relies on empirically attested causal relations to support higher-level hypotheses. Since it is a realist approach, when a scientific theory passes through the sieve of this inductive structure, it becomes at least closer to the truth. As Ghins (p. 92) himself admits, this is still a defense of some kind of best explanation, but he claims that it is different from mere abduction, since “the reason that this explanation is the best is that it is grounded on observations and laws inductively verified a posteriori” (p. 92). For Ghins, this distinction sets his approach apart from typical explanationist strategies, such as Putnam’s “no miracle argument”, rendering it immune to common criticisms that question the truth-conduciveness of the criteria for theory choice. However, I do not perceive a significant difference between Ghins’s account and an abductive appeal to predictive power. The most popular and compelling realist top-down approach is claiming that science works and that we can affect and even impact the world through our so-called best scientific theories, along with their associated metaphysical assumptions.2 I will not argue against this position here, but I would like to point out its similarity to Ghins’s bottom-up approach. An advocate of predictive success could also argue that her reasoning is grounded in an inductive empirical basis, specifically, the historical empirical verification of effective interactions with nature resulting from the development of scientific theories. In both cases, there is an implicit abductive step, the inference that this is the best explanation for the conjunction of phenomena we perceive in the world, and both positions can claim that their accounts are empirically validated and grounded in observations and laws verified inductively a posteriori.
Ghins also criticizes the underdetermination thesis, which asserts that we cannot satisfactorily choose between two, or more, incompatible and empirically equivalent theories. According to Ghins, the flaw of this thesis “lies precisely in its banality” (p. 93). He places the burden of proof on proponents of the underdetermination thesis, arguing that antirealists should not only assert that it is possible to find distinct empirically equivalent alternative theories, but should also demonstrate that this possibility is always the case (pp. 93-4). I believe this demand from Ghins is overtly strong, and that the skeptic would partially fulfill the task if she shows that we have this kind of indecision in at least one case. This alone would open the space for the type of questioning needed for the antirealist argument. Indeed, Bueno & Shalkowski (2020, pp. 463-4) present a quintessential example of incompatible yet empirically equivalent theories: the various interpretations of quantum mechanics, each explaining the same phenomena while imposing different ontological commitments.
Ghins also offers a response to another antirealist argument: Laudan’s pessimistic meta-induction, which draws on well-known historical cases where entities once believed to exist were later shown to be nonexistent according to the evolution of theories. Laudan concludes that the unobservable entities endorsed by our current scientific theories could eventually face similar challenges, with some becoming obsolete terms - like the crystalline spheres of ancient astronomy or the ether (p. 95). Ghins presents a very convincing reply to this objection, highlighting that none of the entities on Laudan’s list had empirically attested causal connections to the phenomena they were meant to explain. For instance, no one ever attested that the crystalline spheres of ancient astronomy were causally linked to the movements of planets; when scientists examined planetary motion and its causes, they concluded that other features of reality accounted for the observed motions (p. 96). Ghins concludes that Laudan’s argument primarily works against top-down realist strategies, as these causal links could have been posited abductively but were never empirically and inductively verified (pp. 96-7). However, given my previous objection, I believe Ghins would need to further clarify how the abductive step implicit in his reasoning is meaningfully distinct from other abductive approaches.
Finally, to ensure the coherence of his scientific realism, Ghins must clarify how we can secure the stability of reference of terms across theoretical changes. An antirealist might argue that different theories are incommensurable, as their theoretical frameworks differ so substantially that even the same term could have completely different denotations within each account.3 Kuhn (1996, pp. 101-2), for instance, suggests that the physical referents of terms such as space, time, and mass in Einsteinian space are entirely different from the same terms in Newtonian theory - for example, mass is conserved for Newton, whereas it is convertible with energy for Einstein.4 Ghins holds that a causal-descriptive theory of reference for natural kind terms is enough to solve the problem of stability of reference across evolving theories (p. 102). Using the the discovery of electron as an example, Ghins asserts that the stability of the referent is not endangered by changes in the connotations of terms used to refer to these entities. He notes that such connotations are absent from the extrinsic causal descriptions of these terms (p. 104). In a Kripkean manner, Ghins attributes the naming of entities like electrons to an initial “baptism”, linked to their detection through external properties (in the case of electrons, the luminous phenomena observed in discharge tubes). As scientists subsequently discovered intrinsic properties (such as the negative charge), we continued referring to the same entity, albeit with a more refined understanding as new, essential properties were uncovered. He concludes: “The meanings of referring terms and successive descriptions can differ, as we saw. Yet this does not prevent mutual understanding among scientists” (p. 104). Additionally, he argues that, without a shared identification of the same referent among scientists, “it would have been impossible to discover its intrinsic properties by new experiments and observations” (p. 104).
I believe that proponents of the incommensurability critique could respond that Ghins has not addressed the core of their objection. The criticism concerns not the terms themselves, but the properties attributed to the referents of these terms - properties that are discovered along with scientific practice. For instance, electrons in Thomson’s theory do not present the same properties as electrons in our current quantum models of the atom. Furthermore, one might argue, as Hanson (1958) did, that observations are deeply theory-laden. Ghins defends a form of direct realism in his book, asserting that detected properties are presented to us directly. Since identification of the referent is initially secured by a description of an extrinsic property evidenced by some phenomenon (p. 109), he believes that reference stability is thus ensured. Yet, if someone claims that the theoretical burden in our observations goes beyond what Ghins acknowledges and, hence, that our ability to notice phenomena is less direct, a more fundamental disagreement arises, complicating acceptance of his argument for reference stability.
I will now turn my attention to Ghins’s defense of a dualist metaphysics of causal powers, which is intertwined with his advocacy for scientific realism and forms the foundation of SRLN’s central goal: proposing a bottom-up approach to science.
2. Dualist Metaphysics of Nature
Ghins defends “an ontology of states of affairs, that is to say facts, which are either thick particulars […] or relations […]. Facts are instantiated properties or relations” (pp. 178-9). Additionally, he considers properties to be universals (p. 179), aligning himself with Armstrong (1978) in both respects. To fully understand this aspect of his theory and how it constitutes a dualist view, we must revisit the distinction between categorical and dispositional properties, as well as how this distinction informs Ghins’s position on the laws of nature.
Ghins supports a propositional philosophical conception of scientific laws, in contrast with ontological conceptions. For proponents of the latter, laws are real entities; for Ghins, on the other hand, they are propositions, not elements of reality itself (p. 114). He also resorts to the classical distinction between particulars and universals, asserting that we cannot conceive of thin particulars, i.e., particulars devoid of all properties. “The thin particular is a logical fiction; only thick particulars exist, which are nothing else than instantiations of properties. […] To be epistemically accessible, a particular must instantiate at least one observable property” (p. 117). This conception of particulars aligns with empiricism and traces back to Russell’s account of particulars as bundles of instantiated properties. Ghins highlights that Russell’s proposal has at least two advantages: it is more economical (since it eliminates the need for a substratum), and it aligns with our ordinary perceptual experiences, given that we perceive properties at specific spatiotemporal locations by means of our senses (p. 118). As he summarizes: “To put it concisely, a (thick) particular is an instantiation of properties (which can also form systems or structures) at some space-time location” (p. 118).
The properties that can be instantiated by particulars fall into two kinds: categorical and dispositional properties. Categorical properties are, in principle, detectable, meaning they can be observed directly or indirectly (p. 120). While there may be categorical properties that are currently unobservable - consider properties detectable today that were undetectable a thousand years ago - Ghins argues that we have no reason to believe in their existence, especially within an empiricist framework. Thus, we cannot establish that a property is categorical if and only if it is observable, but we can conclude that if a property is observable in principle, then it is categorical. Moreover, categorical properties are non-modal, in contrast to dispositional properties (p. 120). Geometric and structural properties, such as being spherical or having the arrangement of carbon atoms in a diamond, are the classic examples of categorical properties (p. 121). However, Ghins suggests that the label ‘categorical’ should also apply to all observable properties, including mass and charge (p. 193).
But what about dispositional properties? According to Ghins, a disposition “endows the particular who bears it with the possibility of acquiring other, not previously possessed, properties (and losing some)” (p. 121). A disposition is manifested when these properties are instantiated (p. 121). Examples include solubility and fragility. While these properties are not yet instantiated, they are not observed, obviously; we cannot see the solubility of sugar or the fragility of glass, we can only directly experience when the manifestation of these dispositions instantiates a categorical property (p. 121), such as dissolved sugar in water or broken glass. Thus, these dispositions are modal, and when their manifestations are described by scientific laws, they are natural or causal powers (p. 121). The metaphysics of categorical and dispositional properties gives rise to various positions regarding their nature, including categorical monism, dispositional monism, and dualism. In SRLN, as the title makes clear, the latter position is defended, that is, the claim that in nature there are both categorical and dispositional properties instantiated by (thick) particulars.
Ghins draws our attention to two issues surrounding natural laws, which he describes as one problem with two aspects. Although he initially applies it to the regularist conception of categorical monism, it can be extended to all accounts of natural laws which any satisfactory explanation should address. This is the problem of identification, which can be divided into an epistemic and an ontological difficulty. The epistemic aspect of the problem asks how we can distinguish laws from mere coincidental regularities. In other words, the central epistemic question about laws is how we can differentiate between the fact that all objects made of uranium 235 have a volume of less than one cubic meter (a natural law) and the fact that all gold cubes are smaller than one cubic mile (a mere contingency, with nothing in nature precluding a gold cube from having a volume of one cubic mile) (pp. 127-8). The ontological aspect of the problem of identification asks: what are laws, after all? Ghins responds to this question by asserting that laws of nature are propositions, as previously mentioned. Furthermore, an account of natural laws must address whether they are contingent or necessary, providing sound arguments for these positions. On the contingentist side, the challenge lies in distinguishing laws of nature from mere contingencies; on the necessaritarian side, it is understanding what sets natural laws apart from logical laws.
Categorical monists claim that dispositional properties, or powers, such as solubility and fragility, can be reduced to the categorical level. In other words, they argue that we are able to provide a description of solubility, fragility, etc. using only categorical properties. Ghins distinguishes two currents within categorical monism: regularists and contingentists (pp. 122-3). I will not address this distinction here, but the main conclusion we have to bear in mind is that categorical monism seems not to be a good explanation because the modal component of dispositional properties cannot be adequately reduced to categorical properties alone. Modalities are not observable, and since categorical properties must be observable, it seems impossible to present powers in a purely categorical manner. On the other hand, there are dispositional monists - Ghins mentions Bird and Mumford as examples (p. 123, 145, 186) - who “reject categorical properties and contend that all natural properties are powers” (p. 145). Among the criticisms against dispositional monism is the “always packing, never travelling” objection, which asserts that, if all properties are powers, then the effect of the manifestation of a power is also a power, leading to a reality that consists solely on shifting powers from a particular to another (p. 187; Psillos, 2006, p. 162). The real issue with such entailment from dispositional monism, as Ghins notes, is that since powers are unobservable (which is the clearest feature that distinguishes them from categorical properties), we would never perceive changes in reality; everything would seem completely motionless (p. 187). Ghins concludes that we need both categorical properties and powers in our metaphysics of nature since,
From an epistemic point of view, the presence of a specific power is attested, indirectly by relying on its manifestations and the truthmaker principle, by the intrinsic similarity between, for example, sugar cubes whose dissolution has been observed and those that have not been immersed in a liquid. Such inductive empirical evidence consists, once again, in the observation of categorical properties (pp. 187-8).
The truthmaker principle mentioned in the passage asserts that true propositions must correspond to facts in the world that make them true (p. 48), a form of truth by correspondence. Since we can indirectly detect the manifestation of powers in the world, which must correspond to a fact, we need an ontology of powers as much as we need an ontology of categorical properties that collaborate in our detection of such powers providing inductive empirical evidence.
After establishing this background, I can return to Ghins’s proposal of a dualist metaphysics of nature. Beyond advocating for an ontology of states of affairs, Ghins argues that predicative expressions in nomological statements describing laws of nature refer to categorical properties, which, as we have seen, encompass all observable properties. Furthermore, natural kind terms (such as ‘gas’) are abbreviations referring to a set of properties that are frequently co-instantiated. In certain cases, like that of electrons, only a single property performs this role, although this is not a requirement. Also, natural kind terms can denote potencies or powers, and worlds where specific powers are instantiated underpin the necessary truth of laws and their associated counterfactuals (p. 179). This raises a question: what are the truthmakers (considering that the framework offered in SRLN includes the truthmaker principle as one of its foundations) for propositions involving powers, such as “sugar dissolves in water”? “I submit that our experience of inevitability counts as evidence for the existence of external entities, which are real natural powers” (p. 181). Ultimately, it seems Ghins relies on a form of “intuition” regarding the inevitability of certain processes, such as sugar dissolving in water, to empirically support the existence of powers. He attempts to uphold both an empiricist view and a dualist metaphysics of nature, but to do so, he resorts to a rationale that seems debatable. If we try to frame his argument in non-intuitive terms, however, this aspect of Ghins’s account becomes unclear. What exactly does an “experience of inevitability” mean and entail? If we have no empirical foundations for detecting modalities, how can we empirically support an appeal to inevitability in an empiricist way?
An attempt to address this question appears in a later passage, where Ghins states: “It is because we have repeatedly seen that sugar cubes dissolve in some circumstances that we conclude that it is a law that sugar cubes always and necessarily behave in such a way” (p. 182). Before this passage, he clarifies that the explanation offered is metaphysical, not scientific, unlike Putnam’s approach with the “no miracle argument”. Moreover, Ghins holds that powers are ontologically prior to observations: “they are the ontological foundation of the conditional necessity of observed regularities described by laws and, at the same time, of the truth of counterfactual propositions” (p. 182). However, a new issue arises here, closely tied to Ghins’s defense of scientific realism and his commitment to a bottom-up approach. Some argue that there is a gap between metaphysics and science, especially in methodological terms. Ghins grounds his scientific realism (a philosophical, not scientific, position, as he notes) on the bottom-up method presented in SRLN as an alternative to the popular top-down, or explanationist, justifications. I have already pointed out the question of whether this approach might still be explanationist. Now, a further question arises regarding what grounds the metaphysics of nature defended by Ghins. If this dualist metaphysics of nature is justified by induction (as Ghins appears to intend), he would need to clarify how this is not begging the question. Furthermore, justifying metaphysics in terms of induction is somehow “risky”, since there will be parts of metaphysical theories that cannot be justified by direct experience - maybe Ghins’s appeal to the inevitability of processes and to the truthmaker principle are examples. The remaining alternative is to justify the theory through abduction, which would fall under the scope of Ghins’s own critique.
Another point Ghins addresses is the necessity of laws of nature. While some argue that the statement “electrons attract positive charges” represents a contingent law - suggesting there could, in theory, be circumstances where electrons do not attract positive charges - others, such as Ellis, attribute the necessity of laws to the essence of the thick particulars referred to - in the previous case, the essence of electrons implies that they attract positive charges (p. 152). Ghins disagrees with this essentialist view held by Ellis, though he maintains that laws of nature are indeed necessary. He claims that there is a de re necessity connecting a power and its corresponding manifestation (i.e., a dispositional and a categorical property), and that some states depend necessarily on a previous instantiation of a power (p. 168). “The fact that a particular possesses a power compels it, as a matter of metaphysical necessity, to behave in a certain way” (p. 169). Powers, therefore, “ground the necessity of lawful propositions as well as the truth of counterfactual conditionals” (p. 169). However, this necessity is constrained by the contingent existence of powers (p. 170). “If the power identified by a specific nomological formula is instantiated in some particular, this particular will necessarily manifest the nomological formula if the conditions or provisos mentioned by the law are satisfied” (p. 180). This is an important step, according to Ghins, in order to distinguish the de re necessity attributed to laws of nature from logical necessity; the latter is true or false in all possible worlds, independently of any other contingent facts. One might question the nature of this necessity for laws. Ghins acknowledges that there is still work to be done on this question (pp. 188-9). The essential point, as he emphasizes, is that “[t]he understanding of the notion of de re possibility is based on the observations of what I can and cannot do and what can and cannot happen in our world” (p. 189). Logical and analytical truths, which are also necessary, are purely formal and do not ground real modalities (p. 189). “There is no logical contradiction in my levitating now. But I clearly know that there is no possibility for me to do so” (p. 189).
Another point worth mentioning concerns the connection established in SRLN between powers and categorical properties. Ghins acknowledges that dispositional monists (such as Mumford and Bird) will object his dualist metaphysics does not take a further step it could take, namely, explaining away categorical properties as covert dispositions, i.e., a step towards dispositional monism. For them, no property is genuinely categorical. Ghins replies that, even in cases where some properties are observed through the laws governing dispositions related to those categorical properties, this is an epistemic issue. A body possesses certain observable properties (mass, charge, etc.) regardless of how we detect them (p. 186). As Ghins notes, and as I previously mentioned, many powers are attested by means of categorical properties, as in the case of dissolving sugar. Furthermore, he contends that the necessity involved in laws of nature is contingent: if there were a world where charges of the same sign attracted each other, then powers would be different from the ones in the actual world (p. 187). Two potential questions may arise against this view: (i) What kind of necessity is that which is constrained by contingent factors? (ii) How are powers and categorical properties connected?
Regarding (i), one might question whether the necessity attributed to natural laws truly qualifies as necessity. We typically believe that a necessary truth or falsity has the same truth value across all possible worlds, irrespective of any particular state of affairs. Ghins limits this strong form of necessity to logical and analytical truths and treats laws of nature as necessary in a weaker, world-dependent sense. Therefore, laws of nature are necessary a posteriori and de re, but, if things were different such as in the case of similar charges attracting each other, then these laws would operate differently. But how Ghins’s stance differ from the contingentist approach espoused by Humeans, for example? Perhaps the best response is to argue that the necessity in Ghins’s view lies in the fact that, given a fixed state of affairs and corresponding truthmakers, no different outcomes could arise; sugar will always dissolve in water and same charges will always repel each other. Yet, one might wonder if the contingentist could claim that this is just pushing contingency further back. Ghins might appeal to necessary a posteriori truths in the way Kripke (1980) describes them: if reality were different, then the necessary truth itself would differ, nonetheless it would still hold as necessary. This suggests a kind of relative necessity. However, necessary a posteriori truths seem less de re in the sense Ghins wants for laws of nature. The atomic number of gold would be attached to it, no matter its value; in contrast, laws of nature appear more substantively grounded in physical reality and its categorical properties.
Finally, (ii) underscores a challenge common to all dualist proposals. When we posit that reality is divided into two distinct domains, we must also account for the relationship between them. While it might seem evasive, one could claim that these domains are simply unconnected and merely co-occur, as in the Leibnizian solution to the mind-body problem, where both sides of the dualist equation are always co-instantiated without any causal interaction. To avoid this approach, the causal connections between both domains must be explicated. Ghins contends that there is a close relationship between powers and categorical properties (p. 186), and this connection must be explicated without a complete reduction of one type of property to the other - otherwise, the proposal would cease to be dualist. It is not clear how a modal property contained in a power can produce observational properties at all; we can only provisionally accept Ghins’s claim that both types of properties may be co-instantiated. Yet, I don’t see this as an outright objection; rather, it points to further inquiries that must be addressed to solidify the dualist metaphysics of nature developed in SRLN. While we cannot assume from the outset that such a connection is impossible to establish, it will undoubtedly be a demanding task, as evidenced by the challenges faced by other dualist approaches throughout the history of philosophy.
3. Conclusion
Ghins’s defense of both scientific realism and a dualist metaphysics of nature in SRLN is robust and stimulates a wide of engaging discussions across numerous areas, including philosophy of science, metaphysics, philosophy of language, epistemology, and modal logic. The impressive scope of fields that can be engaged to support or critique his positions throughout the book underscores its richness. Certainly, SRLN will generate diverse debates and contribute to a deeper understanding of science and the nature of reality, even for those not aligned with scientific realism or a dualist metaphysics of nature.
References
- Armstrong, D. 1978. Universals and Scientific Realism. Cambridge: Cambridge University Press.
-
Bueno, O. & Shalkowski, S. 2020. Troubles with Theoretical Virtues: Resisting Theoretical Utility Arguments in Metaphysics. Philosophy and Phenomenological Research, 101(2): 456-469. https://doi.org/10.1111/phpr.12597.
» https://doi.org/https://doi.org/10.1111/phpr.12597 - Feyerabend, P. 1993. Against Method. 3rd Edition. London: Verso.
- Hanson, N. R. 1958. Patterns of Discovery. Cambridge: Cambridge University Press .
- Kripke, S. 1980. Naming and Necessity. Cambridge, MA: Harvard University Press.
- Kuhn, T. S. 1996. The Structure of Scientific Revolutions. 3rd Edition. Chicago: The University of Chicago Press.
-
Psillos, S. 2006. What Do Powers Do When They Are Not Manifested? Philosophy and Phenomenological Research , 72: 157-176. https://doi.org/10.1111/j.1933-1592.2006.tb00494.x.
» https://doi.org/https://doi.org/10.1111/j.1933-1592.2006.tb00494.x - Putnam, H. 1978. Meaning and the Moral Sciences. London: Routledge.
-
van Fraassen, B. 2017. Misdirection and Misconception in the Scientific Realism Debates. In: E. Agazzi (ed.) Varieties of Scientific Realism, pp. 95-108. Cham: Springer. https://doi.org/10.1007/978-3-319-51608-0_5
» https://doi.org/https://doi.org/10.1007/978-3-319-51608-0_5
-
1
For the sake of brevity, all references to SRLN contain only the page number.
-
2
I would like to thank Thiago Augusto, PhD student from the PPGLM-UFRJ, for presenting me a compelling version of this argument in defense of scientific realism during a discussion in a conference.
-
3
Examples of discussions about such incommensurability can be found in Feyerabend (1993) and Kuhn (1996).
-
4
As Ghins (p. 99 [n. 25]) remarks, Kuhn claims that incommensurability does not entail that proponents of different theories cannot communicate with each other at all (Kuhn, 1996, pp. 198-9). Among the good reasons considered by Kuhn (1996, p. 199) for choosing a theory instead of another, he recurs to fruitfulness, simplicity, accuracy, etc. So, Kuhn’s approach also seems to have a top-down component.
